The three-tempo plan is almost certainly what Bruckner had in mind for both the first movement (Kopfsatz) and the finale, and it underlies the great majority of the recordings in the survey. The following descriptions take that fact as a basis, though not for granted.
Tempos in the introduction. The only explicit tempos are for the beginning, Adagio, for the Intro4 theme, at measure 31, which in Nowak is intended to adumbrate the tempo of the beginning of the exposition: Bewegter, im künftigen Allegro-Tempo. (in the Allegro tempo to follow), and for the renewed Adagio at measure 43. Schalk has, instead, Allmählich bewegter (gradually faster). There are a few early conductors who accelerate in this passage, particularly Furtwängler, Zillig, Abendroth, Pflüger, and Burgin, the last playing from the Schalk score. In the introduction, it is possible to do that since the acceleration is cut off at measure 43 by the adagio theme Intro5 (an evolute of Intro3 which never recurs). But in the recapitulation, at letter N, an acceleration, if taken, would lead directly to the loud version of the A theme, and thus should be started at a slower tempo, as seems to be suggested by Schalk in the introduction. This matter will have to be left to the discretion of the conductor, though it must be said that an acceleration here is very exciting. Meanwhile, it should be said that Intro2 should not be significantly faster than Intro1, which of itself should be a true adagio without being too slow. After all, the time signature is cut time (2/2), which suggests a certain amount of energy.
Dynamics in the introduction. Many conductors begin this symphony with inaudible pizzicatos. Pizzicato strings are one of only two instruments which can be played arbitrarily softly; the other is the kettledrum (the timpani). It is unwise to begin the symphony inaudibly. There should be a very perceptible beginning; Heger writes “leise aber deutlich” in his score. One has plenty of time to play loud later. For example, the great chorale, Intro3 at 18, should be brilliant and martial, with plenty of sound from everybody but especially the tuba. And controversially, it should be just that same way at measure 338 in the recapitulation; see the extensive discussion of this point below.
Tempos in the exposition. Zander 2007 seems to be crystallizing on A = C = 69 and B = 50. This A and C tempo avoids the sluggishness of Harnoncourt, for example, without reaching for an A tempo which would be impossible for C as Botstein does. The B theme should move as fast as possible without giving any hint of hurry, but no slower than 50. Heger marks “beschwingt” (exhilarated, transported) at measure 109 where the arco melody begins. In this movement, there is a complication with two C themes, and in Zander 2007 C1 (measure 161) is only a bit faster than B, with plenty of room for a dramatic acceleration to the A tempo for C2 (measure 199). There is an Allmählich belebend (gradually faster) in the Schalk score at measure 189 which expresses this effect, but it is all right to begin the acceleration four measures sooner as Jochum does again and again, very subtly, if one desires.
At measure 130, at a solemn moment in the Gesangsperiode, Schalk places episemas (tenuto marks) over four quarter notes in the first violin part. At the parallel passage in the recapitulation, measure 402, there are no episemas, but observant Schalk-influenced conductors, and some others like Harnoncourt who can hardly be called Schalk-influenced, add them anyway. There are similar markings in the slow movement of the Second, presumably put there by Bruckner, but they should not be used as an excuse to play the passage especially slowly or with excessive nuance.
In the codetta, at measure 217, there is a downward scalewise progression of chords that is a clear Bruckner signature. It occurs quietly in the first movement of the Second and the last movement of the Ninth, and elsewhere in his symphonies and vocal writing. It is possible that this chordal scale is related to the sleep motive from Die Walküre which is explicitly quoted in the Third and the early version of the Fourth, and it is almost certainly the source of the great brass chorale of the finale of the Ninth. (See my paper on the chorales, “Bruckner’s Hymnal”.) This great importance in Bruckner’s oeuvre prompts certain conductors to take a little time here, and Schalk seems to have the same thought with his poco rall. in measure 220. When done subtly, as it usually is, the effect is quite gratifying.
Dynamics in the exposition. The eexample shows the pizzicato chorale of the B theme, clearly demonstrating the excellent voice-leading. The style is clearly that of a hymn, and as mentioned above I have supplied words, written after Bruckner’s death, which fit the meter of the chorale almost exactly, and besides that, are a reasonable expression of Bruckner’s own view of his work. The example should convince one that the accompaniment needs to be thought of as a chorale, as independent as the chorale is in the famous passage in the Third. In particular, the top voice of the chorale should be brought out, even and especially when they are marked ppp. It is easy to play this music too softly or too slowly, but if one does, a lovely melody and a profound effect go unrealized. One can pronounce the words set to the chorale at half note = 50 or 54 to test the tempo.
Tempos in the development. The extra cadence beginning at measure 231 should not be too slow! The movement does not come to rest here, and the recapitulation of the introduction theme Intro1 which immediately follows should be something of a shock. The tempos of Intro1 and Intro2 should be just as at the beginning, and we will later apply this same requirement to the recurrence of Intro3 and Intro4 at the beginning of the recapitulation. Here we see the operation of Bruckner’s preferred analytical scheme, Part I and Part II, with Part II simply being an expanded and developed recapitulation of Part I. As Bruckner matured, he came closer and closer to having a truly flexible Part II, where the recapitulation of the B theme was preceded by a very brief true recapitulation of only the last part of the A theme group. And in the Fifth, all but one of the elements of the introduction are cited in order: Intro1 at 237, Intro2 at 241 and 259, Intro3 at 338, and Intro4 at 347.
Between these elements of the introduction lies a good deal of research into the synergy of A and C, at the brisk tempo characteristic of those themes. The fast part of the development begins at measure 261, Schalk’s J, where Allegro is written above the staff. Then at measure 283 (Nowak’s L), Schalk writes Etwas langsamer (somewhat more slowly), and at measure 303 (Nowak’s M, Schalk’s L), Noch breiter (yet more broadly). By contrast, Nowak seems to indicate that there is no tempo change necessary here. But non-Schalk conductors, which is to say most people especially younger ones, get a bit slower through this passage just to accommodate the very dense writing at 319. Older Schalk-influenced conductors, like Jochum, Schuricht, Matačić, and perhaps Steinberg, get a lot slower, as does Botstein. I don’t like it much. Perhaps those markings were put into the score in analogy with similar indications in the finale of the Seventh, which definitely do come from Bruckner. But the slow-downs in the Seventh are in the finale, adding grandeur where it is needed, but here in the Fifth, much earlier in the composition, they create a heavy or turgid effect. In his score at measure 303, Heger at first writes “Etwas breiter [somewhat more broadly] (Schalk)”, but later crosses it out and writes “besser dasselbe Tempo beibehalte” [better keep close to the same tempo]. It is typical of him that he would have carefully considered both alternatives, and then felt independent enough of Schalk, whom he must have known in Vienna, to make his own decisions.
In Schalk, there is a fermata at measure 325, just before the obligatory short mention of the B theme which begins in the brass, and another fermata at measure 331, Schalk’s M, just before the B theme in detached woodwind chords and pizzicato strings. These fermatas are a really good idea and could be thought of by anyone just on the basis of the music. In these detached B-theme chords near the end of the development, at 333½ there is often a speedup agreeing with Schalk’s poco accel., and at 335½ a slowdown agreeing with Schalk’s poco rall. This is done by nearly everyone and might also be somewhat automatic. But there is no indication in any score to get slow for the B theme citation. Schalk says to get slower for the B theme, in the exposition at measure 101 with Langsamer (slower), and in the recapitulation at measure 381 with Wieder langsamer (still slower); at 325, all he says is sehr ruhig (meaning very quiet and slow), which Bruckner occasionally uses as a combined indication of tempo and dynamic. So we should not be surprised to find no indication here in Nowak.
Tempos in the recapitulation. The recapitulation may be taken to begin with the statement of the first-movement brass chorale at measure 338 (Intro3). Here there is no indication for a slow tempo in either Nowak or Schalk, though on musical grounds it is very difficult to imagine that theme at anything but a slow tempo. I believe that the generally-incomplete state of editing of this work, combined with Schalk’s inconsistencies, give ample room for the nearly-universal practice of taking measure 338 at the same tempo as measure 18, that is, in the region of 32 or so. Harnoncourt’s tempo of 49 at this point will be difficult for most people, especially considering that his A theme tempo is barely faster. Abbado 1998 also has a fast tempo here for the chorale, and his pizzicato citation of B at 333½ is almost as fast as his A tempo, sounding rather matter-of-fact.
The theme I have called Intro2 occurs with an explicitly-marked Adagio tempo in the Nowak score at measure 15 and 23, and again, Adagio, at measures 241 and 259, but it also serves as the motor for the development at the Allegro tempo, beginning at measure 283 with its latter half expanding into a brilliant cascade of counterpoint at measure 303, as well as in the coda or “second development”. But that use of Intro2 does not make it appropriate to take Intro3 at a similarly fast tempo at 338. In Mus.Hs. 36693, Bruckner’s notation of a tempo at 338 cancels the riten. at measure 335. At the most this notation means that the tempo of Intro3 is the same as that of the B theme, measure 101 or measure 325. It is not a sufficiently strong piece of evidence to justify an allegro tempo anywhere in this passage. Dohnányi’s 1991 performance, with its tempo for measure 338 being 50, a bit faster than his B theme tempo, shows the effect of this kind of thought. Rather than being martial and stirring, it merely seems perfunctory. It impairs an otherwise logical and sensitive performance with a wonderful tearaway coda.
Should the B theme be slower in the recapitulation than in the exposition? Some of the older conductors seem to think so, in a way making up for the fact that the Gesangsperiode is only 44 measures long in the recapitulation compared with 60 measures in the exposition. Indeed there is a sense of Schalk’s Wieder langsamer which could be used to justify taking the theme slower the second time through. But I see no need for that, and also there is the potential of destroying the subtle unity of the movement which must be assiduously developed in the listener’s ear.
Tempos in the coda. In the Schalk version, a fast tempo for the coda requires that the acceleration marked at 437, intended for C2 at 441, be carried through continuously, through the beginning of the coda at 453, to Beschleunigtes Zeitmaass (accelerated tempo) for the peroration at 493. There is no special marking at the beginning of the coda itself. Most conductors like to get fast here, sometimes very fast (Furtwängler, Eugen Jochum). Fast is good, but clarity is good too, and all the details of the writing, which involves the ornate rhythm of Intro2, should be clearly audible at whatever pace is taken. Heger writes very stern instructions for himself at measure 453: “l’istesso tempo / (nicht zu rasch! an den Schluß des Satzes denken! keine Stretta!”. That is, “The same tempo, not too fast! Think of the end of the movement! No tightening up!” And at measure 493, the peroration: “Hauptzeitmaß / l’istesso tempo,” or “Principal tempo, the same tempo.” Yet in the recording he takes a quite fast tempo at that very measure 453, presumably holding that very piece of paper in his hand, which he had not previously reached in the movement. Of course this could represent the cumulative effect, on one man over many years, of many conductors like Jochum who liked to conduct the coda at a speed not previously encountered in the movement. And he did not get as fast as Jochum did.
Dynamics in the coda. Every once in a while, someone tries to apply some degree of nuance to the great forte-fortissimos of Bruckner’s perorations. For example, in an otherwise superb performance of the Fifth dating from September 5, 1998, at measure 497, four measures into the peroration, Abbado suddenly drops back to an mf, and then has a final crescendo to 501 where the violins go up an octave. The coda already has quite a bit of detail and dynamic variety, and it doesn’t need this extra fillip. Schalk, for all the changes he imposed on this piece, says nothing of the sort. He writes ff at 493 and ff sempre at 497 and thus is much closer to Bruckner’s intentions, although few will give him credit for it.